Reforming the labour market for Australian teachers.
labour force development
salary wage differentials
teacher supply and demand
Although there is a general consensus about the importance of teachers to student learning, there is little discussion of the process by which teachers are employed by schools: the teacher labour market. In this paper we analyse whether current institutional features of teacher labour markets are hindering improvements in the quality of teaching in Australia with reference to the empirical literature, such as it exists. We argue, based on mix of a priori and inductive reasoning, that entrenched inflexibilities in the payment system contribute to chronic shortages of qualified teachers in specialised areas and poor incentives for excellent teachers to remain teaching.
We maintain that increased flexibility for schools to remunerate their staff according to whether they possess scarce skills or are outstanding performers will: (a) encourage more and superior graduates to train in the teaching sub-disciplines where there are shortages; and (b) enable schools to retain good instructors and superior youth mentors as teachers.
The following section discusses how teacher labour markets operate and the way in which institutions govern demand and supply for teachers. Section three argues that labour markets for teachers operate inefficiently, as evidenced by chronic shortages in some disciplines and attrition of experienced teachers. Section four examines evidence on what attracts people to teaching and retains them in teaching. The final section canvasses policy options which are likely to relieve shortages and encourage exceptional performance by working teachers.
How teacher labour markets operate
A labour market comprises sellers of labour, in this case qualified teachers, and buyers, that is, school systems and school principals. A labour market can be characterised by the factors that drive demand and determine supply. The dominant factors governing demand for teaching services are the number of people of school age, educational retention rates, the ratio of students to teachers and the cost of hiring teachers. The effective current labour supply of qualified teachers includes all people who have recognised teacher qualifications or currently work or have worked as teachers in Australia. (1) Supply of qualified teachers is limited by institutions: university faculties determine the number of teacher training positions offered and the education authorities decide policies on employing teachers from overseas. Given these parameters, the actual number of supply of teachers available for employment is governed by the pecuniary and non-pecuniary rewards to teaching which attract people to a teaching career. Pecuniary rewards are, of course, wages and probable wages growth. Non-pecuniary rewards are the conditions of employment: the number contact weeks and hours per week, the security of employment, the ease of moving in and out of employment, access to professional development, and the provision of auxiliary staff and teacher aides to ease the work load.
The unrelated nature of these two sets of factors mean that demand will only equal supply by coincidence. Two forces that can balance labour markets are the relative rewards from working (earnings and conditions of employment) and, job opportunities for quahfied teachers in alternative markets. (2)
In Australia, teachers are a large component of the labour market. In 2002 there were 255,000 teachers comprising three per cent of the employed labour force (Australian Bureau of Statistics, 2002). However, teachers do not constitute a single labour market. Teacher labour markets are segmented according to level of schooling--primary or secondary--and within the secondary school labour market, by subject area. Markets are segmented when the type of skills supplied differ so substantially that one type of labour cannot be substituted for another without a considerable loss of productivity. A primary school teacher is very unlikely to be a substitute in a secondary school for a qualified secondary school teacher. In secondary schools, an English teacher cannot substitute a senior mathematics teacher.
Teaching labour markets are not segmented by school sector. A qualified science teacher with two years' experience in the government sector is equally capable of teaching in the non-government sector. Nor are teaching labour markets segmented by employment contract. Teachers within a teacher labour market may be employed on a casual, fixed-term or tenured basis. While they differ in their expected job duration and effective hourly wage rate, they provide the same services.
Table 1 presents 2001 data on the labour force status of qualified teachers. It shows that there is a close correspondence between qualification and level of schooling and that unemployment among qualified teachers is low. Furthermore, high proportions of qualified teachers are not in the labour market, which probably reflects women who leave teaching, either permanently or temporarily, to have children. Of more concern is the considerable loss of qualified teachers from the profession. About twenty-seven per cent of all qualified teachers were employed in occupations other than teaching. However, teachers are not unusual in this regard, Thomas (1988), for example, found that teachers were about middle of the range in terms of graduates working outside their home occupation.
According to the data on teacher outflows presented in Table 2, about six per cent of both primary school teachers and secondary school teachers leave teaching each year. More than one in five of those leave teaching to work in another occupation. The remainder stop working, presumably for family reasons or retirement. Although the percentage moving to other occupations appears small, it is important to note that they are cumulative. Interestingly, the inflow to teaching from other occupations is larger than the outflows. The annual inflow into primary school teaching from other occupations was two per cent and three per cent for secondary school teaching (see Table 3).
Although the annual gross flows from teaching to other labour markets is only 1.6 and 1.4 per cent respectively for secondary and primary teachers, as noted above, over a quarter of those with teaching qualification are working in occupations other than teaching. Increasing supply could involve decreasing the outflow to other occupations and increasing the inflow from other occupations. This can be done by manipulating the incentives to enter or remain in teaching, such as increasing the pecuniary and non-pecuniary rewards. The primary gross flow rates may be less easy to change by direct pecuniary incentives than by non-pecuniary factors, such as flexibility in the hours of work, availability of job sharing and employer-based child care and superannuation.
Do teacher labour markets operate optimally?
This section discusses how effectively teacher labour markets operate. There are two clear indicators of inefficient operation: (a) chronic teacher shortages, and (b) excessive attrition of able teachers to other occupations.
A teacher shortage exists when employers cannot fill the desired number of positions from the start of the school year with appropriately qualified teaching staff at the current wage. A shortage may be caused by insufficient training places, explicit or implicit limits on wages paid, or poor and inaccurate information provided to potential applicants.
State governments expend considerable resources across Australia to regularly assess whether teaching markets are or will be in shortage, surplus or balance. In most states, market demand and supply questionnaires are administered to principals on a regular basis. Unfortunately, many of the questions used do not provide consistent and accurate measures of shortages, and are usually limited to the government school sector.
In addition, statistical models are used to produce forward projections of teacher demand and supply, based on demographic, education and labour force data. Most jurisdictions provide separate forecasts for primary and secondary schools and some (South Australia, Western Australia, Queensland, New South Wales) disaggregate by subject area. In addition, NSW further disaggregates by region. Queensland monitors vacancies every fortnight and provides one-year and ten-year forecasts. In SA and Tasmania, projections from a forecasting model are modified by expert opinion from placement teams. The WA forecasts are for a five-year horizon and NSW uses a seven-year horizon. Victoria conducts two regular surveys of government schools for this purpose. Federally, the Department of Employment and Workplace Relations produces forecasts for the Conference of Education Systems Chief Executive Officers Working Party on Demand and Supply for Teachers (Ministerial Council on Education, Employment, Training and Youth Affairs, 2003; Preston, 2000).
Most of these studies define a shortage as a small number of applications for job vacancies. This is not a shortage, however, in the economists' sense, but indicates that the market is close to balance. If there are no suitable applications for a position, this constitutes a shortage. Evidence that schools are increasing their search activity does not mean there is a shortage but may indicate that the market is close to balance. More appropriate indicators of a teacher shortage are limits on the number of school students who can take a subject because of a lack of teachers, reduced student contact hours, employing teachers to teach subjects they are not fully qualified for, excessive use of relief teachers, and above normal class sizes. Having a teacher in front of every class does not necessarily mean there are no shortages.
Few studies report these indices; nonetheless, those that exist suggest there are chronic shortages in secondary-school science (physics and chemistry), mathematics, information technology and technology subjects, especially in non-metropolitan areas. For example, Ministerial Council on Education, Employment, Training and Youth Affairs (MCEETYA) (2003, p. 58) reports that thirty per cent of Year Twelve mathematics and twenty per cent of science teachers have not completed mathematics or science, respectively, to at least the third year of university. In addition, twenty-eight per cent of Year Eight mathematics teachers were not qualified in mathematics or mathematical education.
Attrition of the most able teachers
An effective labour market should not only attract and retain the required mix of teachers by discipline, but also the most able and best performing teachers. Anecdotally, there is concern that schools often lose their best teachers to more highly-remunerated occupations. There is no Australian evidence on the characteristics of teachers that leave teaching, but United States' studies indicate that high-ability teachers (as measured by Scholastic Aptitude Test (SAT) and National Teacher Exam scores) are the most likely to leave (Henke, Chen & Geis, 2000; Murnane, Singer, Willett, Kemple & Olsen, 1991; Schlecty &Vance, 1981; Weaver, 1983). In addition, physics and chemistry teachers, especially if low paid, were more likely to leave than other secondary-school teachers (Murnane et al., 1991).
To summarise, there is consistent evidence of a shortage of secondary mathematics and science teachers in Australia, especially in non-metropolitan areas. There has not been any systematic attempt to measure the calibre of teachers who leave the profession before retirement, but overseas studies find that more academic and mathematics and science teachers, tend to have higher rates of attrition than other teachers.
Evidence on what attracts people to teaching (and keeps them there)
In order to develop policies that enhance the attractiveness of teaching, we need to understand the motivations of teachers and potential teachers. Non-monetary factors also influence career decisions and there is no reason to suggest that teachers differ from other groups. Many of the non-pecuniary motives are either part of the individual's intrinsic personal preferences or relate to characteristics of the student body over which school administrators and policy makers have little control. For example, studies from Australia, the United States and the United Kingdom have identified both plans for family formation (Bempah, 1994; Dolton, Tremayne & Chung, 2003), the desire to work with children (Bradley, 1983; Committee for the Review of Teaching and Teacher Education (CRTTE), 2003; Milanowski, 2003; Tusin, 1999) and altruistic motives (Ben-Shem & Avi-Itzah, 1991; Young, 1995) as factors effecting the decision to enter the teaching profession. Once trained, teachers' choice of school was influenced, by the characteristics of the potential students: family income, race and academic ability (Hanushek, Kain & Rivkin, 1999), home ownership (Bempah, 1994), and the leadership style of the school administrator (Bempah, 1994).
However, these factors are not open to policy manipulation. Policymakers' scope for attracting and retaining the most able teachers is generally bruited to pay and conditions. Accordingly, it is important to understand how sensitive teacher labour markets are to changes in pay and conditions, especially for mathematics and science teachers. There are very few studies that have directly examined the effects of conditions on teacher attraction. One such example is a US study by Ingersoll (2001) who found poor pay, inadequate support from the school administration, student discipline problems and limited input into school decision-making contribute to higher rates of teacher turnover. By contrast, the role of earnings has been more thoroughly studied. Most of these overseas studies have found that recruitment and retention are quite sensitive to the variations in pay (Dolton, 1990; Dolton & Makepeace, 1993; Dolton & Mavromaras, 1994; Dolton & van der Klaauw, 1995, 1999; Dolton et al., 2003; Gritz & Theobald, 1996; Hanushek et al., 1999; Manski, 1987; Milanowski, 2003; Murnane & Olsen, 1990; Murnane et al., 1991; Zabalza, Turnbull & Williams, 1979). However, this finding is not universal. A recent paper by Fritjers, Shields and Wheatley-Price (2004) found that teachers' decisions to quit were quite insensitive to pay.
Different types of teachers have differing sensitivities to pay variations. Zabalza et al. (1979) found that the supply of male teachers in the UK was quite sensitive to wage variation: a ten per cent increase in wages lead to between twenty-four to thirty-nine per cent increase in supply, while the supply of female teachers was less sensitive. Although men and women were similarly affected by starting salaries, the retention of male teachers was much more responsive to wage growth. In the US, Hanushek et al. (1999) found that a ten per cent rise in starting salaries was associated with a two per cent fall in attrition of new teachers. Salaries have also been found to affect the quality of teachers (Figlio, 2002; Murnane & Olsen, 1989, 1990). Findings from focus group sessions in the US have indicated that mathematics and science college graduates tend to be more motivated by earnings and placed less value the non-pecuniary rewards of teaching, compared with humanities and social science graduates (Milanowski, 2003). If this is generally true, then allowing principals to pay them a premium may alleviate the shortage of mathematics and science teachers.
Another way to evaluate the adequacy of earnings is to examine what the graduate could earn in comparable occupations (Goldhaber & Player, 2003). Both Stinebrickner (2001) and Goldhaber and Lui (2003) found that once qualified, teachers with higher academic ability scores were less likely to enter teaching and more likely to leave teaching for other occupations since the incomes were higher. Hanushek et al. (1999) found that districts with higher salaries tended to recruit teachers with better test scores (5). Therefore, higher salaries are likely to attract and retain a larger number of high ability teachers.
There does not appear to be any analytic studies of the labour market motivations of teachers in Australia although there are data on teachers' stated preferences, relative earnings and job satisfaction. Of the factors that would assist retention, MCEETYA (CRTTE, 2003:92), for example, listed remuneration above reduced workloads and improved employment conditions. The problem may not be with the starting salaries. The Graduate Careers Council of Australia (2000), as demonstrated in Table 9, also shows that between 1977 and 1999 the median starting salaries of education graduates was usually above the overall graduate salary. Furthermore, in general, Australian teachers are not poorly paid. According to Australian Bureau of Statistics data shown in Table 4, teacher's earnings overall are not significantly lower than that for comparable professionals. If it is assumed that teachers have only four weeks of annual leave, then their average hourly earning rate was the same as that for other employee professionals. However, if it is assumed that teachers have twelve weeks of annual leave, then teachers were paid about twenty per cent above the hourly rate for other professional employees with only information technology professionals earning higher rates.
The apparently higher wages of teachers may be because they are, on average, both older and more experienced than workers in other professions and age and experience is associated with higher skills and thus earnings. To assess how important these factors were, we estimated wage equations for the main professional occupational groups using data from the Household, Income and Labour Market Dynamics in Australia (HILDA) survey, 2001 (Wave 1). We found that after adjusting for age, tenure, experience and educational attainment, and other personal and job-related characteristics, school teachers had lower hourly earnings compared to other occupational groups if they were assumed to have four weeks of annual leave per year (7) (Table 5). Given their endowments, school teachers have lower hourly earnings than all other professional occupations except nurses (no significant difference), and social, arts and miscellaneous professionals (a disparate group that includes, among others, lawyers, social workers, religious ministers, photographers, musicians and artists). However, if it assumed that teachers had twelve weeks of annual leave, and accordingly worked seventeen per cent fewer hours, then the pay differential would be very much in favour of teachers.
The HILDA survey also measures job satisfaction (Table 6). Compared with other employees, especially 'other professional' employees, school teachers are highly satisfied with the security of their job (only health professionals score higher) but relatively dissatisfied with the hours they work and how their working hour arrangements impact on their non-work commitments. When all aspects of the job are considered, however, the mean level of satisfaction for school teachers is higher than that for most other professions, although (except for nurses) the differences are not statistically significant. This result was confirmed by regression analysis (not shown here).When job satisfaction was regressed against a wide range of personal characteristics (including age, sex, marital status, educational qualifications and health) as well as occupation, teachers were not any more or less satisfied than workers in other occupational groups.
In sum, the pay of Australian teachers compared with that for other professionals is very sensitive to the assumptions about annual leave. If we assume that teachers have twelve weeks per year (as is often the case) then teachers have higher rates of pay compared to other professionals even after controlling for age and experience. However if it is assumed they have four weeks of annual leave then they are one of the lower-paid professional groups. However, teachers are relatively satisfied with their job.
Regardless of how attractive the teaching profession may be, institutional limitations on the number of teacher training positions are crucial to the supply of teachers. Recently, there has been excess demand for teacher training places. In Victoria in 2001, only about half of all first round applicants to teaching were accepted (Table 7). Interestingly, there was greater demand for primary than for secondary teaching. It is not possible to quantify demand by subject area within secondary-school teaching, since most institutions do not have learning area sub-quotas. Nevertheless, unofficial information from the major training institutions suggest that where sub-quotas exist, they are not filled in the major shortage areas of mathematics, science, and information technology (This information derived from La Trobe, Melbourne and Monash Universities).
There is some concern that the higher numbers of teachers entering the profession in the 1960s and 1970s has led to an excess of older teachers now. According to Table 7, the percentage of employed qualified government primary and secondary teachers aged between forty-five and sixty during 2001 was 40.5 and 42.2 per cent respectively. This is higher than the comparables figures of 31.4 and 33.3 per cent in the non-government sector.
The Committee of Inquiry into Education and Training (Williams Committee) (2003) documented a strong positive trend rise in the median age of teachers since the late 1970s. The increasing age of teachers, together with the superannuation scheme which makes it attractive to retire at fifty-five, suggests there will soon be a shortage of teachers. Over the coming decade the net supply of qualified teachers will have to rise, through training rates, immigration or inflow from other occupations, unless there is a decline in student numbers. However, the Committee for the Review of Teaching and Teacher Education (2003) concludes that student numbers are unlikely to increase for the next decade.
The expected high retirement rate of teachers in the near future is not a concern if teacher training institutions increase their quotas according to departmental forecasts. If forecasting models are reasonably accurate for horizons of two to three years, and teacher training institutions can increase training places within this time frame, then long-term planning for the impending shortage of teachers will not occur. The relatively short period of time required to train a graduate as a qualified teacher means that, given the demand for training positions, the system can be very responsive to short-term changes in the labour market. However, this is not the case for shortages in mathematics and science teachers, which are likely to be exacerbated with the retirement of experienced teachers.
On balance, there is no evidence that overall teachers are poorly paid or dissatisfied with their jobs compared with other professions. However, this does not mean that wage structures are optimal. In particular, the chronic shortage of mathematics and science teachers suggests that their wage rates are too low. Overseas studies have, in general, found that teacher labour supply is quite sensitive to the level of wages. Low training rates for mathematics and science teachers are more likely due to the low level of interest in teaching by mathematics and science graduates, rather than institutional limitations on the number of teacher training positions.
Policies to enhance attraction and retention
If we accept that teacher quality is one of the major influences on student achievement, policies that attract and retain the more able teachers will have a positive effect on student outcomes. (11) To be cost effective, it is desirable that policies focus on specific teacher sub-groups. Our previous discussion about persistent shortages suggests that there are few institutional barriers to the training of mathematics and science teachers and accordingly, policies should focus on attracting mathematics and science graduates to teaching.
The second concern, to reduce the attrition of more able teachers, is only possible if school administrators are able to correctly assess individual teacher performance. Ballou and Podgursky (1998) found that despite lower wages for private school teachers in the US, principals were as satisfied with the quality of their teaching staff and were more successful in retaining the best teachers. They believe that this was because the private system offered better teaching resources and supervision, and because principals had more flexibility to vary pay structures and dismiss teachers for poor performance.
The following section discusses two issues especially important in the Australian context: barriers to using wage premiums to attract and retain 'high demand' teachers and difficulties in identifying the 'best' teachers.
There are no legal or quasi-legal reasons in Australia preventing principals from offering higher pay to different types of teachers. The major criteria for higher pay in the government sector is, however, not type but years of experience, service with current employer and additional responsibilities. Similarly, pay in the Catholic sector is determined by experience and qualifications, not subject matter. While principals can make extra payments above this to retain desirable staff, such payments are typically on an ad hoc basis and are difficult to sustain within the existing budgetary structure. Various state Departments of Education, Employment and Training estimate that in government schools only two to five per cent of teachers receive such incentive payments. Independent schools have greater freedom than the government and Catholic sectors to set wages. Some schools pay teachers less than the government schools and some more, depending broadly on the fee levels. Pay is either determined by collective bargaining or individual contracts. While experience is a major consideration for a teacher's level of pay, some independent schools offer higher rates to teachers trained in high-demand disciplines. In both the government and Catholic jurisdictions, there is little recognition that different and unequal market forces govern different teaching areas. In the independent system, there is greater scope for pay variation.
Teacher unions have opposed wage variation by subject, arguing that teachers should be paid the same rate for the same work. However, teaching in one market segment is not the same as teaching in another. They require different skills and acquired sets of knowledge. The fact that different teachers cannot be substituted for one another is evidence that they are not doing the same work. Similarly, the greater demand for teachers in some subject areas but not in others indicatives that the work is not of the same value. The argument that pay should not vary by subject or year level is heavily based on historical precedent rather than incentive structures designed to ensure students receive the best education.
A major difference in the conditions of employment is tenure status. Teachers can be appointed as on-going (tenured), on fixed-term contract or on a casual or relief basis. It is very difficult to retrench or dismiss tenured staff in the government system. In Victoria since 1999, there have been several restrictions on the ability of schools to hire fixed-term contract and casual or relief staff. Fixed-term contracts can only be offered to teachers replacing a tenured staff member on leave, and casual or relief staff can only be used for short-term absences.
The situation is similar in the Catholic system except that principals may also use a fixed-term contract if it is for a designated short-term subject or if declining student numbers mean that the teacher will not be required the following year. Similarly in the independent sector fixed-term contract staff members are generally not paid a wage premium.
These institutional limitations impose rigidities and inefficiencies on schools by limiting their ability to hire short-term staff where they are most appropriate; that is, in subject areas where future demand for a subject is uncertain or where they are unsure of the suitability of a staff member. This means that schools are hampered in their response to changes in student subject choice and their ability to select the best teachers. In addition, the lack of pecuniary rewards to compensate fixed-term contract teachers for not having tenure means that, on average, the quality of fixed-term contract teachers will be lower than tenured teachers. There is no reason why teachers who prefer limited contracts should be offered a lower total employment package than other teachers. Therefore employers are limited in their ability to offer wage premiums according to the type and quality of teaching services.
Identifying high-performing teachers
Identifying high-performing teachers is difficult and there is a considerable body of overseas, but not Australian, literature in this area. Achievement tests have been used to assess teacher and school performance in the US since 2001 and in the UK since 2000. (Burgess, Groxson, Gregg & Propper, 2001). The major concern is how the value added by the contribution of teachers and schools is measured. Student performance cannot simply be compared between schools or classes since differences can be due to the mix of students' social and socio-economic background and abilities. Therefore, studies of the value added by schools should take into account the students' backgrounds or, more often, ability or prior performance. There are two main approaches (Burgess et al., 2001). The first uses regression analysis to isolate the separate effects of student, school and teacher characteristics on student outcomes--usually test scores or school completion rates. The second method, mainly employed by economists, uses regression analysis to estimate education production functions in order to benchmark individual schools against best practice. Even when using these sophisticated understandings of value added, complications arise. These include tutoring, the costs of testing all students, the tendency for teachers to teach to the test, the incentive to hold back students who perform poorly on the initial test and the sensitivity of the test to small random occurrences. Two or three very high- or very low-performing students can substantially influence summary measures of value added.
Since the mid-1990s, teachers and schools in the US may be rewarded or sanctioned according to the level of, or changes in, student performance in reading and mathematics tests. While student scores tended to increase after the implementation of these policies, the effects on broader based state-wide exams and learning areas not covered by the test is more limited (Jacob, 2002; Klein, Hamilton, McCaffrey & Stecher, 2000; Ladd, 1999). (12) There is evidence that teachers respond to these high-powered incentives by increasing the rate at which poorer students are either placed in special programs which exempt them from the test (Deere & Strayer, 2001; Jacob, 2002) or are held down a year (Jacob, 2002). There are also reports of test 'cheating' by schools (Jacob & Levitt, 2002). Furthermore, there is selective evidence that these incentives have caused teachers to overly teach the test at the expense of low-stake subjects like science and social studies (Jacob, 2002; Deere & Strayer, 2001; Klein, et al., 2000).
This is not to say assessing teacher performance using 'objective' testing methods is impossible but that these methods should be used with caution. These tests can identify teachers whose students consistently perform well above expectations.
Performance appraisals are another approach to assessing teacher performance. Many Australian workplaces have introduced annual performance appraisals over recent decades that are based on a mixture of selected criteria and supervisors' assessments. There does not appear to be any reason why this form of evaluation would not also work in schools.
Hanushek (1986) has argued that the research findings are unequivocal: teachers 'differ dramatically in their effectiveness'. If there were a considerable dispersion of the teacher productivity, then the value to students by increasing the retention of, for example, the top five per cent of teachers would be considerable. To achieve this, however, school managers and supervisors must be able to correctly identify the high performers by fair and objective criteria.
To summarise our argument: teacher shortages and attrition can be addressed by offering wage premiums to mathematics and science teachers, allowing fixed-term contract teachers to obtain higher wages and identifying and rewarding highly effective teachers and by improving the pay or working conditions of particular teacher labour markets. There are no regulations preventing this but it would require changes in attitudes among school administrators regarding appropriate wage structures.
There are few Australian studies on the teacher labour market that inform on recruitment and attrition. This is unfortunate given the importance of the quality of the teachers to student outcomes. Nonetheless, the data canvassed here indicates that there is a continual teacher shortage in secondary-school mathematics and science, especially in non-metropolitan areas.
It appears that fewer mathematics and science graduates wish to become teachers than graduates in the humanities and social science. Wage premiums can be used to attract more mathematics and science graduates. Such premiums need to be maintained during middle and later career to retain existing teachers and attract suitably qualified individuals working in other professions. The problem of recruitment and retention of teachers with scare skills will not be overcome by marginal and one-off attraction incentive schemes such as bursaries and advertising campaigns.
Apart from budgetary considerations and industrial relations issues, there is nothing actually stopping principals from offering teachers with scarce skills higher wages. Accordingly, the deregulation of teacher wages is unlikely to solve the problems we are examining. Raising the minimum wages for specific categories of teachers (for example, in science, mathematics, disadvantaged schools or remote locations) and increasing school budgets could break through stagnation in wage setting norms. This would also provide clear and certain signals to science graduates who may be persuaded to study for a Diploma of Education. To ensure that these incentives are signalled to graduates in mathematics and science, these higher salaries and extended career paths should be formalised in wage agreements and as such would apply to incumbent as well as prospective teachers.
Bonuses and performance loadings also have a role to play in keeping the most able and apt instructors in the classroom. The success of such schemes depends on being able to identify the best teachers and while we are not able to give a definitive solution to this complex and difficult problem, we offer two thoughts. First, increasingly, many non-teaching workplaces are introducing performance loadings and these are usually based on the recommendation of supervisors. There is no reason to suppose that teacher labour markets are any different from other labour markets. Secondly, while performance loadings will never be perfect, their introduction is warranted if they make the delivery of educational services to young Australians better than what they would have been without these loadings.
For these changes to be effective, two enabling factors are required. First, school budgets should be increased to allow salary premiums to be paid and secondly, teachers within schools should recognise and accept that in the interests of better educational services for young people, wage differentials need to change.
Reducing teacher shortages in crucial areas and retaining the most able teachers will no doubt improve the educational outcomes of Australian students and few would disagree that this is a worthwhile goal.
The authors would like to thank John Creedy, Lawrence Ingvarson, Paul Jensen and Adrian Beavis and two anonymous referees for thoughtful comments on this paper. This paper uses the HILDA unit record data file which is funded by the Commonwealth Department of Families, Community Services and Indigenous Affairs.
Auditor General Victoria (2001). Teacher work force planning. Melbourne: Victorian Auditor-General's office.
Australian Bureau of Statistics (ABS) (2002). Schools, Australia: Cat. 4221.0. Canberra: Australian Bureau of Statistics.
Australian Bureau of Statistics (ABS) (1997). Survey of Education and Training Experience: Cat. 6278.0. Unpublished data.
Ballou, D. & Podgursky, M. (1997). Teacher pay and teacher quality. Kalamazoo, Mich: W.E. Upjohn Institute for Employment Research.
Ballou, D. & Podgursky, M. (1998). Teacher recruitment and retention in public and private schools. Economics of Education, 17(3), 393-417.
Bempah, E. (1994). An empirical analysis of teacher mobility. Economics of Education Review, 13(1), 69-77.
Burgess, S., Croxson, B., Gregg, P. & Propper, C. (2001). The intricacies of the relationship between pay and performance for teachers: Do teachers respond to performance related pay schemes? CMPO Working paper series. Bristol: University of Bristol.
Burke, G. (1994). Teachers: Employment in the 1980s and 1990s. In Crowther, F., Caldwell, B., Chapman, J., Lakomski, G. & Olgilvie, D. (Eds.) The Workplace in Education: Australian Perspectives. Sydney: Australian Council for Educational Administrators.
Burtless, G. (Ed.). (1996) Does money matter? The effect of school resources on student achievement and adult success. Washington DC: Brookings Institution Press.
Committee of Inquiry into Education and Training (Williams) (1979). Report of the Committee of Inquiry into Education and Training. Canberra: Australian Government Publishing Service.
Crawley, J., Heckman, J. & Vytlacil, E. (1999) On policies to reward the value added by educators. Review of Economics and Statistics, 81(4), 720-727.
Committee for the Review of Teaching and Teacher Education (CRTTE) (2003). Australia's Teachers: Australia's Future: Advancing Innovation, Science Technology and Mathematics. Canberra: Commonwealth of Australia.
Deere, D. & Strayer, W. (2001). Improving TAAS scores result of better accountability for Texas schools. Texas: Office of University Relations, Texas A&M University.
Department of Education, Employment and Training, Victoria (2000). Forecasting Teacher Demand by Subject. Melbourne: Department of Employment, Education and Training.
Dolton, P. (1990). The economics of UK teacher supply: The graduate's decision. Economic Journal, 100(400), 91-104.
Dolton, P. & Makepeace, G. H. (1993). Female labour force participation and the choice of occupation: The supply of teachers, European Economic Review, 37(7), 1393-411.
Dolton, P. & Mavromaras, K. G. (1994). Intergenerational occupational choice comparisons: the case of teachers in the UK, Economic Journal, 104, 841-863.
Dolton, P. & van der Klaauw, W. (1995). Leaving teaching in the UK: A duration analysis, Economic Journal, 105(429), 431-44.
Dolton, P. & van der Klaauw, W. (1999). The turnover of teachers: A competing risks explanation, Review of Economics and Statistics, 81(3), 543-50.
Dolton, P., Tremayne, A. & Chung, T.-P. (2003). The economic cycle and teacher supply. In Attracting, Developing and Retaining Effective Teachers. Paris: OECD.
Frijters, P., Shields, M. A. & Wheatley-Price, S. (2004). To teach or not to teach? Panel data evidence on the quitting decision. IZA Discussion paper series. Bonn: Forschungsinstitut zur Zukunft der Arbeit.
Goldhaber, D. & Lui, A. (2003). Occupational choices and the academic proficiency of the teacher workforce. In Fowler, W. (Ed). Development in School Finance 2000-01 (pp. 53-75). Washington DC: NCES.
Goldhaber, D. & Player, D. (2003). What different benchmarks suggest about how financially attractive it is to teach in public schools. Madison: University of Wisconsin--Madison, Wisconsin Center for Education Research, Consortium for Policy Research in Education.
Graduate Careers Council of Australia (2000). Graduate Starting Salaries. Parkville: Graduate Careers Council of Australia.
Gritz, R. M. & Theobald, N. D. (1996). The effects of school district spending priorities on length of stay in teaching. Journal of Human Resources, 31(3), 477-512.
Hanushek, E. A. & Jorgenson, D. W. (Eds.). (1996). Improving America's schools: The role of incentives. Washington DC: National Academy Press.
Hanushek, E. A. & Pace, R. R. (1995). Who chooses to teach (and why)? Economics and Education Review, 14(2), 101-17.
Hanushek, E. A., Kain, J. F. & Rivkin, S. G. (1998). Teachers, schools, and academic achievement: Working Paper No. 6691. Cambridge, MA: National Bureau of Economic Research (NBER).
Hanushek, E. A., Kain, J. F. & Rivkin, S. G. (1999). Do higher salaries buy better teachers?: Working Paper No. 7082. Cambridge, MA: National Bureau of Economic Research (NBER).
Henke, R. R., Chen, X. & Geis, S. (2000). Progress through the teacher pipeline: 1992-93 college graduates and elementary/secondary school teaching as of 1997. Washington, DC: U.S. Department of Education, National Center for Education Statistics.
Ingersoll, R. M. (1999). Teacher turnover, teacher shortages, and the organization of schools. Seattle, WA: Center for the Study of Teaching and Policy, University of Washington.
Jacob, B. A. (2002). Accountability, incentives and behavior: The impact of high-stakes testing in the Chicago public schools: Working Paper No. 8968. Cambridge, MA: National Bureau of Economic Research (NBER).
Jacob, B. A. & Lefgren, L. (2002). The Impact of teacher Training on Student Achievement: Quasi-Experimental Evidence from School Reform Efforts in Chicago: Working Paper No. 8916. Cambridge, MA: National Bureau of Economic Research (NBER).
Jacob, B. A. & Levitt, S. D. (2003). Rotten apples: An investigation of the prevalence and predictors of teacher cheating: Working Paper No. 9413. Cambridge, MA: National Bureau of Economic Research (NBER).
Klein, S. P., Hamilton, L. S., McCaffrey, D. F. & Stecher, B. M. (2000). What do test scores in Texas tell us? Education Policy Analysis Archives, 8(49).
Ladd, H. F. (1999). The Dallas school accountability and inventive program: An evaluation of its impact on student outcomes, Economics and Education Review, 18, 1-16.
Lamb, S., Dwyer, P. & Wyn, J. (2000). Non-completion of school in Australia: The changing patterns of participation and outcomes (Longitudinal Studies of Australian Youth Research Report No. 16). Melbourne: Australian Council for Educational Research.
Landon, S. (1997). High school enrolment, minimum wages and education spending. Canadian Public Policy, 23(2), 141-63.
Lee, J.-W. & Barro, R. J. (2001). Schooling quality in a cross-section of countries. Economica, 68, 465-488.
Loeb, S. & Page, M. E. (2000). Examining the link between teacher wages and student outcomes: The importance of alternative labor market opportunities and nonpecuniary variation. Review of Economics and Statistics, 82(3), 393-408.
Manski, C. F. (1987). Academic ability, earnings and the decision to become a teacher: evidence from the National Longitudinal Study of the High School Class of 1972. In Wise, D. (Ed.). Public Sector Payrolls. Chicago: University of Chicago Press.
Ministerial Council on Education, Employment, Training and Youth Affairs (2003). Demand and Supply of Primary and Secondary School Teachers in Australia. Canberra: Australian Government Publishing Service.
Milanowski, A. (2003). An exploration of the pay levels need to attract student with mathematics, science and technology skills to a career in K-12 teaching. Education Policy Analysis Archives, 11.
Murnane, R. J. & Olsen, R. J. (1990). The effects of salaries and opportunity costs on length of stay in teaching: Evidence from North Carolina, Journal of Human Resources, 25(1), 106-24.
Murnane, R., Singer, J., Willett, J., Kemple, J. & Olsen, R. (1991). Who will teach? Policies that matter. Cambridge, MA: Harvard University Press.
Preston, B. (2000). Teacher supply and demand to 2005: projections and context. Canberra: Australian Council of Deans of Education.
Scblecty, P. & Vance, V. (1981). Do academically able teachers leave education? The North Carolina case, Phi Delta Kappan, 63, 106-112.
Stecher, B. M. (2002). Consequences of large-scale, high-stakes testing on school and classroom practices. In L. S. Hamilton, B. M. Stecher, & S. P. Klein (Eds.). Making sense of test-based accountability in education (pp. 79-100). Santa Monica, CA: Rand.
Stinebrickner, T. (2001). A dynamic model of teacher labor supply. Journal of Labor Economics, 19(1), 197-230.
Thomas, C. (1988). Separation from Professions and Para-professions: An analysis of the 1981 Census data for selected professions and para-professions. Canberra: Australian Government Publishing Service.
Wayne, A. J. & Youngs, P. (2003). Teacher characteristics and student achievement gains: A review, Review of Educational Research, 73(1), 89-122.
Weaver, T. (1983). America's teacher quality problem: Alternatives for reform. New York: Praeger Publishers.
Young, B. J. (1995). Career plans and work perceptions or pre-service teachers, Teaching and Teacher Education, 11(3), 281-292.
Zabalza, A., Turnbull, P. & Williams, G. (1979). The Economics of Teacher Supply. Cambridge, MA: Cambridge University Press.
University of Melbourne
Australian Council for Educational Research
(1) However, roughly one-third of all people who take jobs in the economy are not actively looking for work but are induced to take employment because of attractive job offers. Accordingly, the concept of labour supply is fuzzy and includes people with differing degrees of desire for teaching employment.
(2) Balance may also be achieved through similar demand side similar factors. Difficulties of procurement and higher prices may drive consumers to other markets.
(3) Refers to highest qualification. Only about seventy-five per cent of people with an education qualification (as one of their three top qualifications) state education as this highest qualification. This is related to age. Ninety-three per cent of twenty to twenty-four year olds who have an education qualification claim it is their highest qualification, compared with sixty-six per cent of fifty to fifty-five year olds (ABS, 1997).
(4) Includes occupation not stated.
(5) They could not exclude the possibility that the link between salaries and teacher attraction was not causal as wealthy districts also had the students who were more attractive to teach. The short length of the panel made it difficult to clearly disentangle these factors.
(6) Teachers hours have been adjusted on the assumption that they work forty weeks per year compared with forty-eight for other occupations.
(7) As is conventional, hourly earnings was specified as a log function and regressed against variables representing sex, age, tenure, occupation experience, educational attainment, marital status, race, country of birth, health, union membership, employment contract status, employment sector (public or private), workplace size and occupation. The adjusted R-squared value obtained was 0.28.
(8) Refers to highest qualification. Only about seventy-five per cent of people with an education qualification (as one of their three top qualifications) state education as this highest qualification. This is related to age. Ninety-three per cent of twenty to twenty-four year olds who have an education qualification claim it is their highest qualification, compared with sixty-six per cent of fifty to fifty-five year olds (ABS, 1997).
(9) Whether a person with an educational qualification in their highest three qualifications cites it as their highest qualification does not appear to vary by whether they were employed in the education sector or not. While seventy-seven per cent of education sector workers qualified in education cited this as their highest qualification, the rate for those not working and employed in other industries was seventy-four per cent and seventy-five per cent respectively (ABS, 1997).
(10) Includes other government sectors and not stated.
(11) We do not address the issue of what determines teacher quality as it is not strictly relevant to the argument of this paper. The applied research on this topic is scarce and largely foreign and the findings are fairly divergent. Some of the differences in results are likely to be due to the measures of student performance, the scope of control variables and the level of schooling. Many of the US studies only examine primary school pupils and we would not expect their results to generalise for all levels of education. Wayne and Youngs (2003) conducted a review of all US studies on the relationship between teacher characteristics and student outcomes and limited their scope to only those studies which met fairly restrictive criteria regarding the inclusion of proper control variables. This limited their field to twenty-one studies. They concluded that student achievement was greater for teachers trained at more prestigious colleges and for teachers who had themselves achieved better professional test scores during their training. The presence of teacher training and the relevance of specific degrees to discipline taught were only important for mathematics. They found less clear evidence for the role of experience and ethnicity. Studies on the effects of in-service teacher training are also disappointing. Jacob and Lefgren (2002) find no significant effect of teacher training on the mathematics and reading performance for primary school pupils in the US, and they report that out of ninety-three other US studies, only twelve show positive effects.
(12) Stecher (2002), however, reviews case studies and argues that the results are mixed and the net effects uncertain.
Elizabeth Webster is director of the Applied Microeconomics program at the Melbourne Institute of Applied Economic and Social Research and associate director of the Intellectual Property Research Institute of Australia at the University of Melbourne, Vic 3010
Mark Wooden is Professorial Fellow and Deputy Director of the Melbourne Institute of Applied Economic and Social Research at the University of Melbourne, and the director of the Household, Income and Labour Market Dynamics in Australia (HILDA) survey.
Gary Marks is a Principal Research Fellow at the Australian Council for Education Research and a Research Associate with the Melbourne Institute at the University of Melbourne, Vic 3010
Table 1 Labour market status of qualified teachers fifteen to fifty-four years, percentage distribution, Australia, 2001 Occupation / Labour Force Status Primary Secondary Teaching school school qualification (3) teachers teachers Other (4) Primary 44.8 0.9 27.2 Secondary 1.9 49.0 27.8 TOTAL 28.7 19.0 27.5 Occupation / Labour Force Status Not in the Teaching labour qualification (3) Unemployed force Total Primary 1.1 25.7 100.0 Secondary 1.3 19.8 100.0 TOTAL 1.2 23.5 100.0 Source: ABS 2001 Census of Population and housing, unpublished data. Table 2 Annual outflow from teaching (average flows, 1998, 2000 and 2002) Current job Teaching Occupation Other last February Primary Secondary occupations Primary teaching 93.4 0.0 1.4 Secondary teaching 0.00 93.4 1.6 Other occupations 0.00 0.0 93.2 Not working last February 0.50 0.3 75.6 TOTAL 1.30 1.3 88.9 Current job Occupation Not last February working Total Primary teaching 5.0 100.0 Secondary teaching 4.9 100.0 Other occupations 6.6 100.0 Not working last February 23.6 100.0 TOTAL 8.4 100.0 Source: ABS Labour Mobility Surveys, 1998, 2000, 2002 unpublished data. Table 2 Annual inflow from teaching (average flows, 1998, 2000 and 2002) Current job Teaching Occupation Other last February Primary Secondary occupations Primary teaching 93.6 0.0 0.0 Secondary teaching 0.0 94.0 0.0 Other occupations 2.0 3.0 90.7 Not working last February 4.3 2.8 9.3 TOTAL 100.0 100.0 100.0 Current job Occupation Not last February working Total Primary teaching 0.8 1.3 Secondary teaching 0.7 1.3 Other occupations 67.9 86.5 Not working last February 30.6 10.9 TOTAL 100.0 100.0 Source. ABS Labour Mobility Surveys, 1998, 2000, 2002 unpublished data. Table 4 Average weekly earnings for full-time adult non-managerial employees, Australia, May 2000 Average earnings weekly Hours ($) earnings per per Occupation ($) week hour School teachers (four weeks' leave) 898.4 36.0 25.0 School teachers (twelve weeks' leave) 898.4 30.0 (6) 29.9 Natural and physical science professionals 1046.2 38.7 27.0 Accountants, auditors and corporate treasurers 818.4 37.8 21.7 Sales, marketing and advertising professionals 936.3 38.1 24.6 Computing professionals 1210.3 38.4 31.5 Miscellaneous business and information professionals 929.9 38.1 24.4 Miscellaneous social professionals 1094.3 37.2 29.4 All professionals 967.3 37.7 25.7 Source: Employee Earning and Hours. ABS cat. 6306.0, Table 13. Table 5 Hourly wage differentials: School teachers compared professional occupations, Australia, 2001 Estimated % of teacher pay, after controlling for age, experience Sample Occupation etc size School teachers (four weeks leave) 100.0 354 School teachers (twelve weeks' leave) 121.0 354 Other education professionals 118.3 146 Nursing professionals 109.6 214 Other health professionals 128.6 129 Science, building and engineering professionals 109.9 167 Business and information professionals 120.4 512 Social, arts and miscellaneous professionals 99.5 319 Source: Household, Income and Labour Market Dynamics in Australia (HILDA) survey (Wave I). Table 6 job satisfaction by occupation: Mean score (0-10 scale), 2001 Satisfaction with: job Work Occupation Pay security itself School teachers 6.87 8.32 7.87 Other educational professionals 6.77 7.41 8.17 Nursing professionals 5.86 8.62 7.32 Other health professionals 7.19 8.59 8.03 Science, building etc professionals 6.60 7.16 7.85 Business, information professionals 6.88 7.29 7.58 Social professionals 7.13 7.86 7.98 Associate professionals 6.72 7.85 7.74 Other occupations 6.58 7.60 7.54 Satisfaction with: Flexibility to balance Hours work and Overall Occupation worked non-work job School teachers 7.06 6.50 7.73 Other educational professionals 7.00 7.44 7.58 Nursing professionals 6.81 6.92 7.18 Other health professionals 7.40 7.34 7.87 Science, building etc professionals 6.96 7.79 7.59 Business, information professionals 7.14 7.61 7.46 Social professionals 7.15 7.69 7.84 Associate professionals 6.96 7.23 7.62 Other occupations 7.14 7.43 7.62 Source: Household, Income and Labour Dynamics in Australia (HILDA) Survey, Wave I. Table 7 First round applications and subsequent enrolments, Victorian teacher training courses, 1998 to 2001 Course specification Number of applicants 1998 1999 2000 2001 Post Graduate Primary 762 937 1088 1498 Secondary 1318 1358 1465 1884 Undergraduate Primary 1807 2057 2412 3046 Secondary 736 834 855 1091 TOTAL 5398 6189 6759 8334 Course specification Number of first year enrolments 1998 1999 2000 2001 Post Graduate Primary 436 518 545 648 Secondary 1066 1170 1216 1289 Undergraduate Primary 1668 1511 1563 1635 Secondary 692 517 519 578 TOTAL 3862 3716 3843 4150 Sources: Auditor General Victoria (2001, p41) Table 8 Percentage of qualified (8) and employed teachers aged 45 to 60 years in 2001 by sector and occupation, Australia Primary Secondary Other school school occupations teachers teachers (9) Total State/ Territory Government 40.5 42.2 58.5 43.2 Non-government sector 31.4 33.3 46.8 34.2 All sector (10) 37.9 39.2 48.6 38.1